Thirty states have enacted laws mandating private insurers to cover services to diagnose and treat autism spectrum disorders (ASD). Historically, private insurers have excluded or provided minimal coverage for ASD. There has been much debate about these laws. Advocates promote their passage as expanding access to specialized therapies. With enhanced coverage, families will have more opportunity and incentive to use private insurance to pay for their children's care. There is uncertainty, however, about how these laws will be implemented. State mandates require private insurers to cover medically necessary services~ however, because of confusion over what constitutes medically necessary treatment for children with ASD, insurers may opt not to cover certain services. Likewise, provider networks may not grow to meet the increased demand for ASD services, especially if reimbursement rates are set at levels that discourage clinicians from entering the market. To provide empirical evidence about the effects of these mandates, we have partnered with OptumHealth, a subsidiary of United Health Group, a large health insurance plan operating in 25 states with autism mandates implemented from 2009 to 2012 and 23 states without autism mandates during this period. We propose to: 1) examine the effects of state autism mandates on the treated prevalence of children with ASD~ 2) examine the effects of state autism mandates on service utilization patterns and provider reimbursement rates for treatments used by children with ASD~ 3) examine the effects of state autism mandates on spending on services used by children with ASD covered under private insurance~ and 4) explore heterogeneity in the implementation of autism mandates in different tates. For Aims 1-3, we will use health insurance claims to conduct difference-in-difference-in-differences analyses to examine the effects of state mandates by comparing outcomes among: 1) children in mandate states enrolled in fully-insured health plans, which are subject to these mandates~ 2) children in mandate states enrolled in self-insured health plans, who are not subject to these mandates~ and 3) children enrolled in both fully- and self-insured health plans in non- mandate states. Using this rigorous quasi-experimental design, we will identify the effects of this major new state-level policy initiative. For Aim 4, we will interview insurance administrators, provider organizations and consumer advocacy organizations to explore potential effects of mandates not easily studied using claims data, including state-level variation in mandate implementation and effects on care provided and paid for both in and outside health plans. Efforts underway in other states to enact similar legislation in the absence of any evidence on the effects of autism mandates make this proposal particularly timely. Early findings also will be relevant to states' design of the essential health benefits requirements of the Affordable Care Act. The results will help inform the broader debate over the use of insurance design as a mechanism for achieving health policy objectives.